1. Field of the Invention
This invention relates to semiconductor processing apparatus and, more particularly, to a method and apparatus of supplying two different processing gases to a semiconductor wafer processing chamber.
2. Brief Description of the Prior Art
Present-day equipment for the semiconductor industry is moving toward single substrate processing because processing chambers can be made smaller and processing can be better controlled. Further, modern semiconductor vacuum processing systems have been developed to carry out more than one processing step on a substrate without removing the substrate from a vacuum environment. The use of such vacuum systems results in a reduced number of particulates that contaminate the surface of the wafer during processing, thereby improving the device yield.
A typical example of a modern CVD processing apparatus is shown in FIG. 1. In this figure, a single substrate reactor 10 is shown to include a top 12, side walls 14 and a lower portion 16 that, together, define a chamber 18 into which a single substrate, such as a silicon wafer 20, can be loaded. The wafer 20 is mounted on a susceptor 22 that can be rotated by a drive 23 to provide a time-averaged environment for the wafer 20 that is cylindrically symmetric.
A preheat ring 24 is supported in the chamber 18 and surrounds the susceptor 22. The wafer 20 and the preheat ring 24 are heated by light from a plurality of high-intensity lamps, schematically indicated as 26, mounted inside of the reactor 10. The top 12 and lower portion 16 of the reactor 10 are typically made from clear quartz which is transparent to the light from lamps 26. Quartz is generally used to make up the top 12 and lower portion 16 because it is transparent to light of both visible and IR frequencies, because it exhibits a relatively high structural strength and because it is chemically stable in the process environment of the chamber.
During the deposition process, processing gas (whether reactant or dopant) is supplied to the interior of the chamber 18 from an exterior source, schematically represented by two tanks 28. The gas flows from the gas supply 28 along a gas supply line 30 and into the chamber 18 via a gas inlet port 32. From the port 32, the gas flows across the preheat ring 24 where it heats up, across the susceptor 22 and wafer 20 in the direction of the arrows 34, to be evacuated from the chamber 18 through evacuation port 36. The dominant shape of the flow profile of the gases is laminar from the gas input port 32 and across the preheat ring 24 and the wafer 20 to the exhaust port 36, even though the rotation of the wafer 20 and thermal gradients caused by the heat from the lamps 26 do affect the flow profile slightly.
The above-described CVD processing chamber can accommodate a number of different processes taking place. Each process differs, depending on the desired end result, and has different considerations associated therewith.
In the polysilicon deposition process, doped or undoped silicon layers are typically deposited onto the wafer using processes such as low-pressure chemical vapor deposition (CVD). In this process, a reactant gas mixture including a source of silicon (such as silane, disilane, dichlorosilane, trichlorosilane or silicon tetrachloride) and, optionally, a dopant gas (such as phosphine, arsine or diborane) is heated and passed over the wafer to deposit a silicon film on its surface. In some instances, a non-reactant carrier gas, such as hydrogen, is also injected into the processing chamber, together with either or both of the reactant or dopant gases. In this process, the crystallographic nature of the deposited silicon depends upon the temperature of deposition. At low reaction temperatures, the deposited silicon is mostly amorphous; when higher deposition temperatures are employed, a mixture of amorphous silicon and polysilicon or polysilicon alone will be deposited.
One problem with the doped polysilicon deposition is that the temperature dependence of dopant incorporation is the opposite of the temperature dependence of the polysilicon deposition rate. This is because adjusting the temperature to obtain thickness uniformity in the polysilicon layer produces a non-uniform dopant incorporation. This is because the dopant gas has, in the past, been incorporated into the processing gas before it is injected into the chamber. There is therefore no control of the dopant gas flow independent of the flow of the silicon species processing gas.
In another process, the nitride deposition process, a stream of reactant gas, which is a mixture of ammonia (NH.sub.3) and any one of the various silane species, is injected into the chamber. These two gases react at room temperature to produce small crystals. In the arrangement shown in FIG. 1, the gas storage 28 is shown to include two tanks, both of which feed into a single supply line 30. If these tanks contained ammonia and silane respectively and the line 30 were at room temperature, this reaction would occur and particles would form along the entire length of the supply line 30 and within the manifold 32. These particles are undesirable, as they are a source of contamination in the chamber 18; and their existence should therefore be eliminated.
In addition, it has been found that some reactant gases pass through the gap between the preheat ring 24 and the susceptor 22. This causes deposition on the back side of the susceptor 22 and on some of the other components in the lower portion of the chamber 18. Such deposition is both wasteful and undesirable, as it requires additional cleaning to remove.
Accordingly, a need has arisen for a system of supplying reactant/dopant gases to a semiconductor processing chamber which overcomes these different problems.